RE R-Mix... What the heck?

Just checked out the vid Freddie posted. Can someone explain this a little further? Am I to assume I can take a stereo track and isolate elements and tweak and/or remove them? Because that would be freaking crazy and super useful and really make X2 about a million times more desirable to me. Cheers.

@John... My gawd... I have literally HUNDREDS of hours of live recordings and rehearsals that would be perfect if it weren't for a couple of wayward tracks and audience/room noise. I could release like twenty albums if this thing actually worked. Oh dude dude dude... NOW I am psyched. Please tell me this is reality.

Depending on how engrained they are in a mix, it looks like it. A vocal with multiple panned harmonies and loads of delays and/or reverb might be more difficult to extract though.

I'm sure you could use to to re-master stuff to a degree or maybe use it for extracting samples from songs.

As to the sound quality... You could see some comb filtering and phasing artefacts potentially. I wouldn't trust the cake videos on sound quality. It's going to take un-biased testing to check that one out.


To me, it looks like a cool tool for a bit of fun, but I wouldn't ever imagine using it when making my own music.

But for maybe analysing in further detail the construction of pro songs, it could be good. As well as taking samples for people who do that sort of music.

But I don't imagine it to be much special. The high and low stuff is just going to be a simple band pass filter. The stereo component is going to be probably a M/S type process with something like Channel Tools and phase cancellation blah blah. I'm sure the maths isn't too complex. With the right routing you can probably just do it with channel tools and an eq. I may of course be wrong here, I'm only speculating.

Will the R-Mix in X2 work on 24bit audio files?









That would be kinda cool.

That will change everything.

fixed

I mentioned this somewhere a couple of days ago and folks said it was generally not well received, and just a glorified karaoke machine. 

you can turn a sausage grinder backwards, but pigs will not come out the top.  I fear this is more bloat rather than a professionally useful tool.  Is X2 gonna support the MCU?

"I mentioned this somewhere a couple of days ago and folks said it was generally not well received, and just a glorified karaoke machine."

That may be because Roland advertising has suggested that karaoke is where it's at.


R-Mix has the potential to be fantastic for film sound where it is routine to wish you had a tool which can remove a small background noise from location dialog.

R-Mix has the potential to rescue a drum track with a squeaky pedal by removing the squeak.

There's all kinds of possibilities... not to mention the potential to dig in to sounds to grab samples that may be a basis for some further sound design.

There are uses such as these that Roland hasn't promoted with much vigor. It just emphasizes using it for karaoke backing tracks and so the result is what it is.


I've been looking forward to a R-Mix that will load 24bit files. The current free standing version only works on 16bit files.

*sad* Oh well... But now that we're on the topic I guess I could use any and all advice on isolating elements in a stereo track so I can attempt to remix the take. IE: Isolate bass, guits, vox, drums. Toss them into a new project. Mix and add/replace what is needed. You guys are talking Melodyne here... is that just for vox or can it work on other stuff too?

Melodyne is an amazing product. There's no doubt. But! Complex operations can be tedious and the results sketchy. It does some things much better than others. DNA blew my mind when it was first introduced but I wouldn't begin to expect it isolate all the different vocal and instrumental elements of a song and do it flawlessly with great results.

@pdlstl... Thanks for that. Really I'd be focusing on the good takes/performances anyway so I wouldn't really need anything TOO crazy. Just enough to even things out properly and get most of the unwanted noise out of the way. Obviously complete reconstruction of a track is daunting at best and is why I got so excited about this. Anyway, I've heard people here refer to Ozone (I think that's the program I am referring to). I know it's expensive but would that be a tool to save up for for this kind of work? If I actually COULD figure out how to do this properly I'm sure I could get a lot of work from the bands I know not to mention how useful something like that would be to me personally. It'd be worth an investment. Cheers.

The R-mix could also be helpful as an analyst by showing where things sit in the mix and help isolate things for a clearer mix.

Guitarpima


The R-mix could also be helpful as an analyst by showing where things sit in the mix and help isolate things for a clearer mix.

To be honest, that was one of things I first thought as I watched the video. Then I realized how much discipline it took me to trust what I hear rather than what I see. So for me, in that application, it would be a bad thing. YM, of course, MV. However, it might just be the best thing since sliced bread. We'll just have to wait and see.

Michael Five


a tool for specialized filtering in raw tracks could be most useful...

So would audiosnap.

pdlstl

Guitarpima


The R-mix could also be helpful as an analyst by showing where things sit in the mix and help isolate things for a clearer mix.

To be honest, that was one of things I first thought as I watched the video. Then I realized how much discipline it took me to trust what I hear rather than what I see. So for me, in that application, it would be a bad thing. YM, of course, MV.

It could assist your ears by letting you hear what is conflicting. I agree though. Trust your ears more than your eyes. 

It's the little blobs of color that threw me off. Couldn't we lasso those little blobs and isolate them to snag what we want and rebuild from there? Then tell the program to say "snag any blobs similar to this" with a sensitivity parameter to allow for variations. Or is this like crazy moon talk?

Now that I think of it why shouldn't this work? I had it backwards in my head though. I'd watch the blobs for the part I was trying to isolate and then do a general clean up to get all the unwanted stuff out of the way. Then a more pinpointed clean up. If the graphical representation of the audio is accurate there shouldn't be too many competing frequencies and even if there is it will match up in the final mix anyway. This might work... maybe... kind of... sort of.